INDEX

Explanations

information related to various technical terms and data entries

oai_token-act-pair · gpt-3.5-turbo

instances of numeric data or identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Entered

-0.75

 Gutenberg

-0.57

Lub

-0.57

 FOIA

-0.52

Pag

-0.52

icter

-0.51

 redundancy

-0.50

 seism

-0.50

Beg

-0.50

 Epstein

-0.50

POSITIVE LOGITS

 respectively

0.83

 alike

0.82

 itself

0.79

cycles

0.69

,,,,

0.68

rect

0.66

Latin

0.65

orea

0.64

wide

0.62

whatever

0.62

Activations Density 0.989%

information related to various technical terms and data entries

instances of numeric data or identifiers

No Comments

No Known Activations