INDEX

Explanations

terms that are used to refer to specific concepts or things

oai_token-act-pair · gpt-3.5-turbo

definitions and explanations of terms

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

idav

-0.80

asive

-0.77

iar

-0.71

 Veter

-0.69

ó

-0.67

JM

-0.67

pour

-0.67

aqu

-0.66

asions

-0.66

©¶æ¥µ

-0.66

POSITIVE LOGITS

bidden

0.80

 qualities

0.77

 abbre

0.74

 initials

0.71

 shorthand

0.69

 plural

0.69

 irregular

0.68

 anything

0.68

vow

0.67

 atomic

0.66

Activations Density 0.126%

terms that are used to refer to specific concepts or things

definitions and explanations of terms

No Comments

No Known Activations