INDEX

Explanations

instances of the term "excessive" followed by a variety of contexts

oai_token-act-pair · gpt-3.5-turbo

references to the term "excessive" in various contexts, particularly relating to force, actions, or conditions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hran

-0.87

Reviewed

-0.83

mberg

-0.79

rol

-0.78

arity

-0.77

stone

-0.77

udder

-0.77

debian

-0.75

bard

-0.74

ingen

-0.73

POSITIVE LOGITS

 amounts

0.99

 punishment

0.94

 excessive

0.93

 accumulation

0.83

 amount

0.81

 expenditure

0.80

 eleph

0.77

 heat

0.75

 restraints

0.74

 delays

0.73

Activations Density 0.013%

instances of the term "excessive" followed by a variety of contexts

references to the term "excessive" in various contexts, particularly relating to force, actions, or conditions

No Comments

No Known Activations