INDEX

Explanations

action verbs indicating physical or mental activity

oai_token-act-pair · gpt-3.5-turbo

verbs indicating actions or capabilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Sabb

-0.64

}"

-0.64

mes

-0.61

Lenin

-0.61

Jur

-0.57

 Completed

-0.57

Cly

-0.57

IL

-0.56

 Mell

-0.55

 meant

-0.55

POSITIVE LOGITS

 safely

0.91

 freely

0.84

ibly

0.83

 muster

0.80

 easily

0.79

uate

0.78

 anytime

0.78

ivably

0.77

 anywhere

0.73

odder

0.71

Activations Density 0.311%

action verbs indicating physical or mental activity

verbs indicating actions or capabilities

No Comments

No Known Activations