INDEX

Explanations

proper nouns or names that end with the letters "off"

oai_token-act-pair · gpt-3.5-turbo

terms related to the concept of being "off" or not active

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

ãĤ©

-0.82

aceous

-0.74

ãĤ§

-0.70

CVE

-0.66

ãĤ£

-0.64

SHARE

-0.64

ãĥ£

-0.63

umbing

-0.62

entary

-0.62

POSITIVE LOGITS

ices

1.07

ense

1.00

erence

0.87

ENSE

0.85

rey

0.84

enses

0.83

sets

0.82

ments

0.82

ishly

0.82

enders

0.81

Activations Density 0.015%

proper nouns or names that end with the letters "off"

terms related to the concept of being "off" or not active

No Comments

No Known Activations