INDEX

Explanations

verbs related to actions or processes

oai_token-act-pair · gpt-3.5-turbo

phrases and terms related to following or adhering to policies and decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

oiler

-0.69

nesday

-0.67

hene

-0.66

iment

-0.65

URI

-0.63

 guiActiveUnfocused

-0.63

jah

-0.62

ãĤ¨ãĥ«

-0.62

ridge

-0.61

lua

-0.61

POSITIVE LOGITS

 various

0.99

 programs

0.94

 numerous

0.87

 these

0.85

 myriad

0.82

 many

0.79

 traditional

0.78

 surveys

0.78

 educational

0.76

 other

0.75

Activations Density 0.766%

verbs related to actions or processes

phrases and terms related to following or adhering to policies and decisions

No Comments

No Known Activations