INDEX

Explanations

words related to the concept of being strongly tied or connected

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "bound" and its variations, often relating to limitations or connections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

issance

-0.75

visors

-0.69

cano

-0.63

PE

-0.60

icago

-0.58

————

-0.57

blems

-0.56

ptive

-0.56

 fielded

-0.56

paces

-0.56

POSITIVE LOGITS

less

1.38

lessly

1.23

unin

0.94

ging

0.94

fold

0.93

lessness

0.90

ing

0.90

binding

0.90

aries

0.87

sym

0.87

Activations Density 0.066%

words related to the concept of being strongly tied or connected

occurrences of the word "bound" and its variations, often relating to limitations or connections

No Comments

No Known Activations