INDEX

Explanations

titles or names with special characters like symbols or non-alphanumeric characters

oai_token-act-pair · gpt-3.5-turbo

references to specific media titles or related concepts

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

anwhile

-0.68

dig

-0.67

ioned

-0.67

 shack

-0.67

 scattering

-0.67

 cyan

-0.67

enta

-0.66

hiba

-0.66

 bleeding

-0.65

lda

-0.64

POSITIVE LOGITS

catentry

0.92

âĹ¼

0.89

Ħ¢

0.87

º

0.86

Mech

0.86

£

0.85

į

0.83

Premium

0.82

Hon

0.80

WER

0.79

Activations Density 0.329%

titles or names with special characters like symbols or non-alphanumeric characters

references to specific media titles or related concepts

No Comments

No Known Activations