INDEX

Explanations

verbs representing desires or preferences

oai_token-act-pair · gpt-3.5-turbo

expressions of desire or expectation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

hement

-0.75

andy

-0.61

haz

-0.59

vati

-0.58

travel

-0.57

hero

-0.56

velt

-0.56

shoot

-0.56

lves

-0.55

stant

-0.55

POSITIVE LOGITS

 happening

0.92

 happen

0.82

 beforehand

0.77

 ABOUT

0.74

"},"

0.74

 about

0.72

 when

0.71

 transpired

0.71

 versus

0.70

 happens

0.70

Activations Density 0.175%

verbs representing desires or preferences

expressions of desire or expectation

No Comments

No Known Activations