INDEX

Explanations

proper names or terms starting with "Aleks"

oai_token-act-pair · gpt-3.5-turbo

references to specific names or entities

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

kefeller

-0.70

eele

-0.66

artifacts

-0.65

 endlessly

-0.63

 hereafter

-0.63

 bearer

-0.62

jri

-0.62

buster

-0.61

verning

-0.61

 specificity

-0.58

POSITIVE LOGITS

OND

0.71

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.69

henko

0.64

 Glac

0.64

uria

0.63

rical

0.63

phia

0.61

Skill

0.61

abeth

0.61

 Pipeline

0.60

Activations Density 0.176%

proper names or terms starting with "Aleks"

references to specific names or entities

No Comments

No Known Activations