INDEX

Explanations

instances of the word "talking" and its variations, indicating a focus on dialogue or discussions

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Features by Cosine Similarity

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

Plots

Explanation

Show Test Field

Default Test Text

IFrame

Link

Not in Any Lists

Negative Logits

emale

-0.76

iverpool

-0.75

uilt

-0.73

feeding

-0.70

boa

-0.70

peria

-0.68

metic

-0.67

eele

-0.65

cffff

-0.64

proc

-0.63

POSITIVE LOGITS

 Heads

0.87

 louder

0.86

 about

0.82

 aloud

0.81

 Points

0.78

 loudly

0.78

heads

0.74

 voices

0.72

Points

0.72

 filibuster

0.71

Activations Density 0.020%

instances of the word "talking" and its variations, indicating a focus on dialogue or discussions

No Activations Uploaded