INDEX

Explanations

phrases with the prefix "mis-" that denote incorrect or faulty actions

oai_token-act-pair · gpt-3.5-turbo

words related to misrepresentation or distortion

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hetti

-0.76

Â¯Â¯Â¯Â¯

-0.74

uesday

-0.68

 Enhancement

-0.65

 Skydragon

-0.65

 Parties

-0.65

loo

-0.65

 Dynamics

-0.64

 Norn

-0.64

ILA

-0.62

POSITIVE LOGITS

cellaneous

1.11

appropri

1.08

pell

1.01

pelled

0.94

beh

0.87

informed

0.87

jud

0.86

aligned

0.86

ouri

0.81

per

0.80

Activations Density 0.012%

phrases with the prefix "mis-" that denote incorrect or faulty actions

words related to misrepresentation or distortion

No Comments

No Known Activations