INDEX

Explanations

phrases urging the reader to verify or explore additional content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 rendition

-0.80

Attempt

-0.68

Church

-0.67

 suscept

-0.64

ãĥĨãĤ£

-0.61

è£ħ

-0.61

Imagine

-0.60

CONCLUS

-0.59

chief

-0.59

nown

-0.59

POSITIVE LOGITS

lists

0.99

mate

0.98

boxes

0.91

list

0.84

ers

0.84

ups

0.80

hee

0.79

points

0.75

ering

0.74

box

0.74

Activations Density 0.012%

phrases urging the reader to verify or explore additional content

No Comments

No Known Activations