INDEX

Explanations

words related to prize or honor categories

oai_token-act-pair · gpt-3.5-turbo

verbal forms related to categorization and classification

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

recy

-0.72

erest

-0.67

ishops

-0.65

put

-0.65

rises

-0.65

Origin

-0.63

sav

-0.60

razil

-0.59

istry

-0.59

 Rossi

-0.59

POSITIVE LOGITS

TAIN

0.87

ãĥī

0.86

iser

0.86

iatus

0.86

ãĥķãĤ¡

0.82

owitz

0.77

otonin

0.75

Ø©

0.69

hoe

0.67

Â¯Â¯Â¯Â¯

0.67

Activations Density 0.015%

words related to prize or honor categories

verbal forms related to categorization and classification

No Comments

No Known Activations