INDEX

Explanations

interviews and reviews

oai_token-act-pair · gpt-3.5-turbo

key terms related to media and entertainment content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

theless

-0.90

sic

-0.76

maximum

-0.72

olson

-0.71

hap

-0.69

orthy

-0.68

fortunately

-0.68

 lowly

-0.68

omever

-0.68

perature

-0.68

POSITIVE LOGITS

 Scenes

0.97

 Manager

0.96

 Tracker

0.94

 Dates

0.93

 Rankings

0.92

 Tracks

0.92

 Plans

0.92

 Crimes

0.91

 Reviews

0.90

 Theme

0.90

Activations Density 0.335%

interviews and reviews

key terms related to media and entertainment content

No Comments

No Known Activations