INDEX

Explanations

words related to rankings, importance, and influence

oai_token-act-pair · gpt-3.5-turbo

references to significant entities, events, or concepts in various sectors including politics, sports, and culture

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

phasis

-0.78

erity

-0.75

Laf

-0.70

 Santos

-0.69

displayText

-0.68

iris

-0.68

isson

-0.67

iless

-0.66

ritz

-0.65

é¾į

-0.65

POSITIVE LOGITS

 EVER

0.91

 achie

0.88

 mammal

0.83

 accomplishment

0.82

 contender

0.81

 negotiator

0.79

 imaginable

0.78

Ever

0.78

 asset

0.77

 politician

0.77

Activations Density 0.371%

words related to rankings, importance, and influence

references to significant entities, events, or concepts in various sectors including politics, sports, and culture

No Comments

No Known Activations