INDEX

Explanations

The main thing this neuron does is find mentions of territories being annexed

oai_token-act-pair · gpt-3.5-turbo

terms related to territorial acquisition or annexation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

pid

-0.73

orah

-0.71

pton

-0.70

acter

-0.69

ukemia

-0.67

DAY

-0.64

$$

-0.59

xious

-0.59

 fres

-0.58

life

-0.57

POSITIVE LOGITS

 annexed

1.13

 annexation

1.06

 annex

0.88

itized

0.85

ãĥ¼ãĥĨãĤ£

0.84

ificant

0.84

imates

0.84

éĹĺ

0.80

ãģĨ

0.78

hers

0.77

Activations Density 0.013%

The main thing this neuron does is find mentions of territories being annexed

terms related to territorial acquisition or annexation

No Comments

No Known Activations