INDEX

Explanations

statements related to logic, decisions, and agreements

oai_token-act-pair · gpt-3.5-turbo

logical coherence and rational connections in text

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

 civilisation

-0.75

 organisations

-0.64

\",

-0.60

Firstly

-0.60

âĦ¢

-0.60

 organisation

-0.59

 Â®

-0.59

CLAIM

-0.58

\":

-0.57

CHAPTER

-0.57

POSITIVE LOGITS

 likewise

0.85

itto

0.84

same

0.76

 similarly

0.75

oother

0.71

 unspecified

0.69

trak

0.68

 unsuccessfully

0.65

ospons

0.65

TBA

0.65

Activations Density 1.607%

statements related to logic, decisions, and agreements

logical coherence and rational connections in text

No Comments

No Known Activations