INDEX

Explanations

mentions of the word "Mother"

oai_token-act-pair · gpt-3.5-turbo

references to "Mother" in various contexts

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

NRS

-0.76

ript

-0.74

EY

-0.73

ickr

-0.70

ramer

-0.70

aping

-0.70

RAFT

-0.67

okers

-0.66

ype

-0.65

REP

-0.64

POSITIVE LOGITS

 Mother

0.97

 Teresa

0.94

ship

0.85

Mother

0.82

hood

0.80

Joy

0.75

hesis

0.74

 Anne

0.73

 Mary

0.73

heses

0.73

Activations Density 0.018%

mentions of the word "Mother"

references to "Mother" in various contexts

No Comments

No Known Activations