INDEX

Explanations

proper nouns related to geographical locations and organization names

oai_token-act-pair · gpt-3.5-turbo

references to peaks and notable locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 rabb

-0.75

 trustee

-0.74

 rabbi

-0.71

itiz

-0.69

 rust

-0.66

afort

-0.64

tab

-0.64

ãĤ¶

-0.63

oa

-0.63

 dign

-0.62

POSITIVE LOGITS

 Peak

3.77

peak

3.35

 Peaks

1.92

 peak

1.10

 Highlands

1.07

 peaks

1.02

 Maximum

0.97

 Wilderness

0.96

Pe

0.92

max

0.91

Activations Density 0.039%

proper nouns related to geographical locations and organization names

references to peaks and notable locations

No Comments

No Known Activations