INDEX

Explanations

proper nouns, specifically human names or titles

oai_token-act-pair · gpt-3.5-turbo

phrases or expressions that include the word "of" related to names

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

iasm

-0.74

 touch

-0.71

agy

-0.69

issions

-0.67

ists

-0.66

 Divide

-0.65

idelines

-0.65

estyles

-0.65

ptions

-0.62

venants

-0.60

POSITIVE LOGITS

 Uzbek

0.80

 culprit

0.74

ãĤ·ãĥ£

0.68

ãĤ¬

0.65

 Shots

0.64

ãĤ´

0.63

 suspects

0.62

Ox

0.62

otta

0.62

 Killer

0.61

Activations Density 0.128%

proper nouns, specifically human names or titles

phrases or expressions that include the word "of" related to names

No Comments

No Known Activations