INDEX

Explanations

mentions of radio personalities and activities related to radio shows

oai_token-act-pair · gpt-3.5-turbo

mentions of radio-related content

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

rir

-0.83

jamin

-0.79

 Bane

-0.79

ortium

-0.78

inho

-0.75

laus

-0.75

 Snape

-0.74

tall

-0.72

rin

-0.70

 Arbor

-0.70

POSITIVE LOGITS

activity

1.09

 stations

1.09

actively

1.08

 broadcasts

1.04

 station

1.00

 broadcasters

0.98

active

0.97

 radio

0.94

 antennas

0.92

 announcer

0.90

Activations Density 0.011%

mentions of radio personalities and activities related to radio shows

mentions of radio-related content

No Comments

No Known Activations