INDEX

Explanations

words related to making specific requests or demands

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "requested" and its variations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

phal

-0.74

smoking

-0.73

bows

-0.73

tan

-0.71

tra

-0.70

bow

-0.69

VERTISEMENT

-0.68

sam

-0.67

pson

-0.67

cour

-0.67

POSITIVE LOGITS

 permission

0.85

 requested

0.83

 requests

0.78

 Parenthood

0.77

 anonymity

0.73

 subpoen

0.71

REAM

0.71

erville

0.69

 authorization

0.67

 accommodation

0.67

Activations Density 0.027%

words related to making specific requests or demands

occurrences of the word "requested" and its variations

No Comments

No Known Activations