INDEX

Explanations

pronouns or nouns referring to people

oai_token-act-pair · gpt-3.5-turbo

the relative pronoun "who," indicating a focus on subjects or individuals 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

Quantity

-0.74

AN

-0.62

======

-0.62

arity

-0.60

ANK

-0.59

DIT

-0.58

dit

-0.58

UNE

-0.56

ENN

-0.56

mud

-0.56

POSITIVE LOGITS

hesda

0.76

 comprises

0.74

 accompanies

0.73

pires

0.71

©¶æ

0.69

 rocked

0.69

 appealed

0.68

 comprise

0.67

 redes

0.67

 benefited

0.66

Activations Density 0.293%

pronouns or nouns referring to people

the relative pronoun "who," indicating a focus on subjects or individuals in various contexts

No Comments

No Known Activations