INDEX

Explanations

verbs related to starting or initiating an action

oai_token-act-pair · gpt-3.5-turbo

instances of uncertainty or qualifications in statements

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

peria

-0.68

 sidx

-0.63

 retali

-0.62

LEASE

-0.61

ban

-0.56

ben

-0.55

ordial

-0.55

ERG

-0.54

roid

-0.54

Plan

-0.54

POSITIVE LOGITS

 admittedly

0.76

olls

0.71

mittedly

0.71

 nutshell

0.68

 frankly

0.67

 incidentally

0.65

 paraph

0.64

Anyway

0.64

*)

0.64

versely

0.64

Activations Density 0.420%

verbs related to starting or initiating an action

instances of uncertainty or qualifications in statements

No Comments

No Known Activations