INDEX

Explanations

phrases related to providing systems or solutions

oai_token-act-pair · gpt-3.5-turbo

terms related to behavioral studies or scientific analyses of processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

iatus

-0.93

ackets

-0.75

obiles

-0.71

 delinquent

-0.70

merce

-0.67

agen

-0.67

 hiatus

-0.65

 depl

-0.62

 postp

-0.61

lled

-0.61

POSITIVE LOGITS

helps

0.85

 integrates

0.78

albeit

0.78

 automate

0.74

 integrate

0.72

 synergy

0.72

 improves

0.71

 helps

0.71

 antioxid

0.68

Improve

0.68

Activations Density 0.870%

phrases related to providing systems or solutions

terms related to behavioral studies or scientific analyses of processes

No Comments

No Known Activations