INDEX

Explanations

mentions of a prominent or main position within a context

oai_token-act-pair · gpt-3.5-turbo

instances of the word "lead" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

aves

-0.72

cale

-0.70

ramid

-0.67

afort

-0.66

 Remem

-0.63

ikes

-0.63

iannopoulos

-0.63

 symmetry

-0.62

terday

-0.62

urches

-0.61

POSITIVE LOGITS

better

0.88

boards

0.77

 negotiator

0.76

lead

0.76

gers

0.70

ership

0.70

 poisoning

0.69

 blocker

0.66

 coffin

0.66

 counselor

0.65

Activations Density 0.028%

mentions of a prominent or main position within a context

instances of the word "lead" in various contexts

No Comments

No Known Activations