INDEX

Explanations

words related to specific names or locations, particularly those starting with "Wy"

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or entities, particularly names that start with "Wy"

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

inates

-0.93

ãĥ¼ãĥĨ

-0.81

å£«

-0.77

å§«

-0.74

istically

-0.73

orage

-0.69

inators

-0.69

raints

-0.68

ividual

-0.67

ocaust

-0.67

POSITIVE LOGITS

atts

1.09

nesday

1.04

ank

0.87

tch

0.87

nec

0.86

cz

0.85

izard

0.82

vern

0.82

combe

0.82

rote

0.81

Activations Density 0.024%

words related to specific names or locations, particularly those starting with "Wy"

references to specific individuals or entities, particularly names that start with "Wy"

No Comments

No Known Activations