INDEX

Explanations

names or words related to the author "Poilievre"

oai_token-act-pair · gpt-3.5-turbo

words and terms related to various topics and possibly proper names

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

¥µ

-0.84

 tremend

-0.79

Zen

-0.77

ItemThumbnailImage

-0.74

ãĤ¤ãĥĪ

-0.69

 admiration

-0.66

 affinity

-0.66

ADRA

-0.63

 scoreboard

-0.62

Gameplay

-0.61

POSITIVE LOGITS

fecture

0.84

nder

0.83

anski

0.83

eworthy

0.83

ciples

0.82

ulent

0.75

theless

0.75

incial

0.73

omial

0.72

quit

0.71

Activations Density 0.061%

names or words related to the author "Poilievre"

words and terms related to various topics and possibly proper names

No Comments

No Known Activations