INDEX

Explanations

names and terms related to the location "Cap"

oai_token-act-pair · gpt-3.5-turbo

references to "Cap" or related terms, likely indicating locations or characters associated with "Cap."

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

ĪĴ

-0.96

hower

-0.90

 silence

-0.84

perse

-0.77

¿½

-0.74

 anten

-0.72

td

-0.71

WAYS

-0.67

cause

-0.67

Reloaded

-0.66

POSITIVE LOGITS

itol

1.20

itals

1.18

acity

1.09

illary

1.05

aic

1.04

abilities

1.04

rice

0.98

rices

0.97

aign

0.96

illus

0.95

Activations Density 0.010%

names and terms related to the location "Cap"

references to "Cap" or related terms, likely indicating locations or characters associated with "Cap."

No Comments

No Known Activations