INDEX

Explanations

proper nouns, possibly related to politics or specific locations

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

senal

-0.75

sonian

-0.73

gger

-0.72

steen

-0.71

WT

-0.71

WARE

-0.70

\\\\\\\\\\\\\\\\

-0.70

BALL

-0.69

berry

-0.68

chy

-0.68

POSITIVE LOGITS

idential

1.09

assion

1.09

iction

1.06

inct

1.05

ributed

1.05

ractions

1.04

ression

1.03

irmed

1.02

luence

1.01

acts

1.00

Activations Density 0.611%

proper nouns, possibly related to politics or specific locations

No Comments

No Known Activations